Photochemical control of RNA structure by disrupting π-stacking

Article abstract of DOI:10.1021/ja306304w

Photolabile nucleotides that disrupt nucleic acid structure are useful mechanistic probes and can be used as tools for regulating biochemical processes. Previous probes can be limited by the need to incorporate multiple modified nucleotides into oligonucleotides and in kinetic studies by the rate-limiting step in the conversion to the native nucleotide. Photolysis of aryl sulfide 1 produces high yields of 5-methyluridine, and product formation is complete in less than a microsecond. Aryl sulfide 1 prevents RNA hairpin formation and complete folding of the preQ₁ class I riboswitch. Proper folding is achieved in each instance upon photolysis at 350 nm. Aryl sulfide 1 is a novel tool for modulating RNA structure, and formation of 5-methyluridine within a radical cage suggests that it will be useful in kinetic studies.

Full text of DOI:10.1021/ja306304w

Communication
pubs.acs.org/JACS
Photochemical Control of RNA Structure by Disrupting π‑Stacking
Marino J. E. Resendiz,† Arne Schon,‡ Ernesto Freire,‡ and Marc M. Greenberg*,†
̈
†Department of Chemistry and ‡Department of Biology, Johns Hopkins University, 3400 North Charles Street, Baltimore, Maryland
21218, United States
S
* Supporting Information
nucleotide via a radical pair that undergoes disproportionation
within a solvent cage.
ABSTRACT: Photolabile nucleotides that disrupt nucleic
acid structure are useful mechanistic probes and can be
used as tools for regulating biochemical processes.
Previous probes can be limited by the need to incorporate
multiple modified nucleotides into oligonucleotides and in
kinetic studies by the rate-limiting step in the conversion
to the native nucleotide. Photolysis of aryl sulfide 1
produces high yields of 5-methyluridine, and product
formation is complete in less than a microsecond. Aryl
sulfide 1 prevents RNA hairpin formation and complete
folding of the preQ1 class I riboswitch. Proper folding is
achieved in each instance upon photolysis at 350 nm. Aryl
sulfide 1 is a novel tool for modulating RNA structure, and
formation of 5-methyluridine within a radical cage suggests
that it will be useful in kinetic studies.
When designing 1 we took advantage of the precedent for
electron-rich aryl sulfides undergoing carbon−sulfur bond
homolysis upon irradiation within the same region of the
aforementioned o-nitrobenzyl derivatives.15,16 The methyl
group was included to ensure that a substituent in the
dihydropyrimidine would adopt a pseudoaxial orientation thus
ensuring base stacking disruption. Molecular mechanics energy
minimization indicated that the 5R-diastereomer of 1 would
adopt a conformation in which the aryl sulfide adopts a
pseudoaxial position while maintaining the C3′-endo con-
formation of the ribose ring that is present in A-form RNA
(Figure 1). This is consistent with the expected larger A-value
ucleic acids, particularly RNA, undergo conformational
1−3
Nchanges in their secondary and tertiary structure.
Complex structural changes occur during RNA folding, which
in the case of a riboswitch is triggered by a ligand.4 There is
continuing interest in developing methods for detecting and
controlling nucleic acid folding in order to provide a better
mechanistic understanding of these processes.5,6 In addition,
methods for controlling nucleic acid hybridization and structure
are also under development as tools for regulating gene
expression and other biochemical processes.7−9 Photochemistry
is a popular tool for triggering nucleic acid structural changes
because it provides spatiotemporal control and can be used in
conjunction with time-resolved detection methods. One
approach utilizes a modified nucleotide(s) in which a
photolabile group alters the nucleobase’s ability to form base
pairs thus destabilizing the biopolymer’s folded structure.10,11
Ultraviolet irradiation restores the native nucleotide’s structure.
The o-nitrobenzyl photoredox reaction is often employed in
these systems. This venerable photochemical reaction proceeds
in high chemical yields and is very useful in preparative nucleic
acid chemistry.12 However, the need in some instances to
incorporate multiple photolabile nucleotides in a single
oligonucleotide to impart structural control is a limitation.13
Furthermore, time-resolved studies are limited by the time scale
(up to 1 min) on which the leaving group (e.g., the native
nucleotide) is released from an intermediate formed following
irradiation.14 We report a new photolabile molecule that
disrupts nucleic acid structure by perturbing base stacking. This
molecule is amenable to use in applications that require a
shorter time scale because its photolysis produces a native
Figure 1. Energy minimized (Spartan) structure of photochemical
substrate 1. Two perspectives are shown to highlight different
structural aspects of 1.
of the methyl group and calculations on related dihydropyr-
imidines.17 The 5R-diastereomer was examined because this
was expected to be the major isomer upon sulfenylation (3) of
the enolate of 2 (Scheme 1).16,18 Indeed, a single isomer of 1
was obtained following deprotection of 3, which was difficult to
purify, and was carried on to the requisite phosphoramidite (4).
Photolysis (350 nm) of an aqueous solution of 1 (4 min)
under aerobic conditions produced 5-methyluridine in 78%
yield (85% mass balance). The yield of 5-methyluridine (MeU)
was unaffected by β-mercaptoethanol (BME) at concentrations
as high as 3.5 M, and increasing the BME to 7 M (50% by
volume) only decreased the yield to 65%.21 The inability of the
thiol to compete with the disproportionation reaction suggests
that 5-methyluridine is predominantly formed within a radical
cage (Scheme 2). Importantly, if one assumes that BME reacts
with the alkyl radical with a rate constant of at least 2.6 × 106
Received: June 28, 2012
Published: July 24, 2012
© 2012 American Chemical Society
12478
dx.doi.org/10.1021/ja306304w | J. Am. Chem. Soc. 2012, 134, 12478−12481

Journal of the American Chemical Society
Communication
a
The effect of 1 on secondary structure was analyzed using
CD spectroscopy. Molecule 5 displayed all of the spectroscopic
features at 25 °C that are characteristic of an RNA hairpin
(Figure 2A).22,23 These include intense bands displaying
Scheme 1
a
Key: (a) H2, Rh/Al2O3; (b) TDBMSCl, (c) LDA, disulfide; (d) TFA;
(e) DMTCl; (f) TBDMSCl, AgNO3, pyridine; (g) Phosphitylation.
Scheme 2
M−1 s−1,19 the lack of an effect of 3.5 M BME on the product
yield indicates that 5-methyluridine is formed in less than a
microsecond. This is several orders of magnitude more rapid
than product release from the o-nitrobenzyl photochemical
reaction and suggests that 1 could be used for probing
conformational changes that occur on the microsecond time
scale.14
Figure 2. CD spectra of RNA oligonucleotides (65 μM) 5 and 6. (A)
6 at 25 °C before and after hυ overlaid with 5. (B) Comparable spectra
for 7 and 8.
positive (λmax = 272 nm) and negative (λmax = 212 nm)
ellipticities and two weaker bands at (−) 240 and (+) 225 nm.
Heating the sample resulted in hypochromicity of both of the
signature bands at 272 and 212 nm, the analysis of which
yielded a Tm value of 43.1 0.8 °C.21 In contrast, CD spectra
of 6, which contains photolabile 1 instead of MeU, lacked the
features that correspond to A-form RNA (Figure 2A), most
notably the band at 212 nm. However, photolysis (20 min) of 6
yielded a product that exhibited a CD spectrum that contains
the same bands as that obtained from 5 (Figure 2A). Further
support for the successful photochemically induced trans-
formation of 6 into 5 is gleaned from the Tm (44.2 1.1 °C) of
the material produced upon photolysis of 6, which is within
The ability of 1 to disrupt secondary structure was initially
examined in RNA hairpins (5−8) (Scheme 3). The hairpin
Scheme 3
experimental error of that measured for 5 (43.1
0.8).21
Comparing the CD spectra of 8 before and after photolysis
with that of 7 shows the same overall effect (Figure 2B). Hence,
a single molecule of aryl sulfide 1 disrupts hairpin formation but
is readily converted to MeU, which promptly folds to yield CD
spectra containing the same characteristic features as those
produced from independently synthesized oligonucleotides
containing MeU (5, 7).
Successful modulation of secondary structure by 1 led us to
examine its ability to control folding in a more complicated,
biologically relevant molecule. The preQ1 class I riboswitch
specifically recognizes 7-aminomethyl-7-deazaguanine (preQ1),
which is an intermediate in queuosine biosynthesis. The
aptamer region of this riboswitch contains only 34 nucleotides,
and its folded structure from different species is well
characterized.24−28 In the absence of preQ1 the receptor from
Fusobacterium nucleatum contains a (5 bp) stem-loop region
(Scheme 4).24,27 This stem-loop is retained upon preQ1
sequences were chosen based upon a previous report in which
the described systems could serve as a starting point such that
incorporation of 1 would turn off hairpin formation.20 Aryl
sulfide 1 was introduced into the oligonucleotides (6, 8) via
standard solid phase synthesis methods using slight mod-
ifications for the coupling of 4 and subsequent deprotection of
the biopolymers.21 Oxidation was carried out using t-BuOOH
in order to minimize sulfoxide and/or sulfide formation. The
oligonucleotides containing 1 or MeU were separable from one
another by denaturing and native polyacrylamide gel electro-
phoresis.21 Photolysis for 10 min completely converted the
modified oligonucleotides into ones that comigrated with the
otherwise identical RNAs containing MeU.
12479
dx.doi.org/10.1021/ja306304w | J. Am. Chem. Soc. 2012, 134, 12478−12481

Journal of the American Chemical Society
Communication
The lack of an effect of preQ1 on the CD spectra of RNA
molecules containing 1 in place of either U4 (12)21 or U32 (13,
Figure 3B) indicates that the dihydropyrimidine prevents
pseudoknot but not stem-loop formation. However, the
dihydropyrimidine significantly destabilizes the stem region.
In the absence of preQ1 the (UV-melting) Tm of 12 was 12 °C
Scheme 4
lower (46.2
0.3 °C) than that of 10 (58.5
0.3 °C).
Photochemical conversion of 1 to MeU, which is complete by
gel electrophoresis within 6 min, is accompanied by the
anticipated change in 13’s CD spectrum upon pseudoknot
formation in the presence of preQ1 (Figure 3B).21
In summary, we have developed an unnatural nucleotide (1)
that modulates secondary and tertiary nucleic acid structure.
The molecule is converted in high yield to 5-methyluridine
upon photolysis via a process that occurs in less than a
microsecond. We anticipate that 1 will be a useful tool in
studying RNA folding and in developing photoregulated
nucleic acid molecules.
binding, which also gives rise to pseudoknot formation. We
examined the ability of 1 to modulate folding of the preQ1
aptamer by substituting the dihydropyrimidine (in separate
experiments) for uridines that are involved in pseudoknot (U32)
and stem formation (U4).
Before examining the impact of 1 on the aptamer we
investigated the effect of MeU substitution on folding. No
difference was detected in the CD spectra of 9−11, indicating
that substituting MeU for uridine does not alter aptamer
folding (Figure 3).21,29 For instance, in the absence of preQ1
ASSOCIATED CONTENT
■
S
* Supporting Information
Experimental procedures. Mass spectra of all oligonucleotides.
Sample autoradiograms showing photochemical conversion of
1 in oligonucleotides. CD spectra. This material is available free
of charge via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
■
Corresponding Author
mgreenberg@jhu.edu
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We are grateful for generous financial support from the
National Institute of General Medical Sciences (GM-054996 to
M.M.G. and GM57144 and GM56550 to E.F.). E.F. thanks the
National Science Foundation for support (MCB-1157506).
M.J.E.R. thanks the NIGMS for a Research Supplement to
Promote Diversity in Health-Related Research. We thank an
anonymous reviewer for bringing ref 29 to our attention.
REFERENCES
■
(1) Behrouzi, R.; Roh, J. H.; Kilburn, D.; Briber, R. M.; Woodson, S.
A. Cell 2012, 149, 348−357.
(2) Woodson, S. A. Curr. Opin. Chem. Biol. 2008, 12, 667−673.
(3) Sclavi, B.; Sullivan, M.; Chance, M. R.; Brenowitz, M.; Woodson,
S. A. Science 1998, 279, 1940−1943.
Figure 3. CD spectroscopy analysis of the effect of 1 on preQ1 folding.
(A) 11 with and without preQ1 (3 equiv). (B) 13 with and without
preQ1 (3 equiv) before hυ and with preQ1 after hυ. [RNA
oligonucleotides] = 10 μM. Comparable spectra for 9, 10, and 12
are in the Supporting Information.
(4) Haller, A.; Souliere, M. F.; Micura, R. Acc. Chem. Res. 2011, 44,
1339−1348.
(5) Kreutz, C.; Kaehlig, H.; Konrat, R.; Micura, R. J. Am. Chem. Soc.
2005, 127, 11558−11559.
(6) Hobartner, C.; Mittendorfer, H.; Breuker, K.; Micura, R. Angew.
̈
Chem., Int. Ed. 2004, 43, 3922−3925.
the CD spectrum of 11 is consistent with hairpin loop
formation (Figure 3A). Addition of preQ1 results in an increase
in the intensity of the main bands, which is consistent with
formation of a pseudoknot.30,31 A slight bathochromic shift was
also observed upon the addition of preQ1 in each instance.
(7) Tang, X.; Maegawa, S.; Weinberg, E. S.; Dmochowski, I. J. J. Am.
Chem. Soc. 2007, 129, 11000−11001.
(8) Tang, X.; Dmochowski, I. J. Angew. Chem., Int. Ed. 2006, 45,
3523−3526.
̌ ́ ́
(9) Kielkowski, P.; Macíckova-Cahova, H.; Pohl, R.; Hocek, M.
Angew. Chem., Int. Ed. 2011, 50, 8727−8730.
(10) Wenter, P.; Furtig, B.; Hainard, A.; Schwalbe, H.; Pitsch, S.
̈
Angew. Chem., Int. Ed. 2005, 44, 2600−2603.
(11) Hobartner, C.; Silverman, S. K. Angew. Chem., Int. Ed. 2005, 44,
̈
7305−7309.
12480
dx.doi.org/10.1021/ja306304w | J. Am. Chem. Soc. 2012, 134, 12478−12481

Journal of the American Chemical Society
Communication
(12) McGall, G. H.; Barone, A. D.; Diggelmann, M.; Fodor, S. P. A.;
Gentalen, E.; Ngo, N. J. Am. Chem. Soc. 1997, 119, 5081−5090.
(13) Young, D. D.; Lively, M. O.; Deiters, A. J. Am. Chem. Soc. 2010,
132, 6183−6193.
(14) Il’ichev, Y. V.; Schworer, M. A.; Wirz, J. J. Am. Chem. Soc. 2004,
̈
126, 4581−4595.
(15) Fleming, S. A.; Jensen, A. W. J. Org. Chem. 1996, 61, 7040−
7044.
(16) San Pedro, J. M. N.; Greenberg, M. M. Org. Lett. 2012, 14,
2866−2869.
(17) Miaskiewicz, K.; Miller, J.; Ornstein, R.; Osman, R. Biopolymers
1995, 35, 113−124.
(18) Barvian, M. R.; Greenberg, M. M. J. Org. Chem. 1993, 58, 6151−
4.
(19) Newman, C. A.; Resendiz, M. J. E.; Sczepanski, J. T.; Greenberg,
M. M. J. Org. Chem. 2009, 74, 7007−7012.
(20) Molinaro, M.; Tinoco, I., Jr. Nucleic Acids Res. 1995, 23, 3056−
3063.
(21) See Supporting Information.
(22) Williams, D. J.; Hall, K., B. Biochemistry 1996, 35, 14665−14670.
(23) Hannoush, R. N.; Damha, M. J. J. Am. Chem. Soc. 2001, 123,
12368−12374.
(24) Rieder, U.; Kreutz, C.; Micura, R. Proc. Natl. Acad. Sci. U.S.A.
2010, 107, 10804−10809.
(25) Kang, M.; Peterson, R.; Feigon, J. Mol. Cell 2009, 33, 784−790.
(26) Klein, D. J.; Edwards, T. E.; Ferre-D’Amare, A. R. Nat. Struct.
Mol. Biol. 2009, 16, 343−344.
(27) Rieder, U.; Lang, K.; Kreutz, C.; Polacek, N.; Micura, R.
ChemBioChem 2009, 10, 1141−1144.
(28) Spitale, R. C.; Torelli, A. T.; Krucinska, J.; Bandarian, V.;
Wedekind, J. E. J. Biol. Chem. 2009, 284, 11012−11016.
(29) All experiments were carried out at low concentrations where
dimerization is not expected to contribute significantly. See: Santner,
T.; Rieder, U.; Kreutz, C.; Micura, R. J. Am. Chem. Soc., DOI: 10.1021/
ja3049964.
(30) Johnson, K. H.; Gray, D. M. J. Biomol. Struct. Dyn. 1992, 9, 733−
745.
(31) Klepper, F.; Polborn, K.; Carell, T. Helv. Chim. Acta 2005, 88,
2610−2616.
12481
dx.doi.org/10.1021/ja306304w | J. Am. Chem. Soc. 2012, 134, 12478−12481